Optimising optical: converter, SFP module and cable choices

Yes - many. And they can all hear an improvement. Here just a couple of quotes:

"But with the FTLX8574D3BCV a wider soundstage, less harshness and improved mid-bass suggest that these modules really reduce noise. At $128/pair, that was the highest ROI audio purchase I’ve made in a long time.

Recently I changed to these TLX1475D3BTL. Result More focus, precision and dynamics, more natural sound.

Cisco SFP-10G-AOC5M series of optical SFP/cables are the absolute best-sounding AOCs of many that I’ve tried."

The HOT ones seems to come from Finisar.

Would have been fun just to do a test. But I don’t think Finisar will work with my Cisco Meraki switch.

Torben

1 Like

I think Cisco OEM SFP+ modules from Finisar in the same way Intel do, so might be worth a look if you want to experiment.

The modules on my router are branded with both logos, Intel and Finisar, so it should be easy to find out.

Well, after a quick look, I wasn’t able to find any Finisar SFP+ branded for Meraki, but this post seems to suggest they might work: Tried some SFPs in Meraki switch.

I’d want to dig a bit more deeply before making a purchase, though!

2 Likes

Torben, Jeremy is spot on, you can indeed try out non-Cisco/Meraki SFPs in Meraki Switches.

This is their official position;

Meraki offers branded SFP modules, and while we do not prevent third-party accessories from functioning, users should conduct their own tests to ensure proper compatibility.

4 Likes

Has arrived today from the US. Until now it “plays” just fine with my Cisco Meraki Go Switch:

Torben

3 Likes

If I’m not mistaken, the Cisco GLC-LH-SMD SFP is the exact same module, made by Finisar in Malaysia, but rebadged Cisco


(See the Serial number “SN: FNS…” thats the give away that its Finisar sourced :laughing:)

4 Likes

Currently I use:

  • Corning SMF 28 ULTRA Single mode (G.657.A1 ) - OS2 single mode grade B fiber patch cable - connector LC UPC to LC UPC duplex - bend insensitive.

and am very happy with it. 2 meters cost 14.72 + shipping.

But now I’ve discovered the new “magic word” :grin: - Zero Water Peak fiber optic cable:

I have now ordered:

  • 1.8 meter - Commscope TERASPEED zero water peak single mode LC/UPC to LC/UPC fiber patch cord, 1.6 mm duplex, LSZH - Comply with G.657.A1 and G.652.D (FFWLCLC42-JXF006)

Cost 36 EURO + shipping

Torben

1 Like

Am I allowed to politely ask why? It’s a modest expense in hifi terms, to say the least, so you could just be an inquisitive kind of guy, but I wondered if you expected a different optical cable to sound different in any particular way.
Enjoy playing, whatever!

I don’t know - if you don’t try out things you never know. If it don’t, than I have the lastest single mode standard - “Full spectrum, zero water peak single mode (ZWP-SM)”.

The cable is compliant with ITU-T G.652.D and ITU-T G.657.A1.

Torben

1 Like

When you say “latest standard”… are these guys wrong to say there isn’t one?

Good on you. I love to experiment myself but I usually have some sort of hypothesis in mind, like I might try thing A because it appears/is claimed it has effect B which might deliver benefit C.

The case made by Fosco for ZWP seems to relate to data loss which increases the distance over which data can be reliably transmitted. I presume you don’t think the cables you use today are giving you data losses. This therefore wouldn’t get through my Is There A Hypothesis To Test Here filter. But I love it when I can learn from someone else’s experimentation!

Report back.

Understanding ITU-T Standards for Various Optical Fibers:

If you don’t like the word standard - just use another word. Fine with me.

I will contact all the big companies (Corning, FT, Commscope …) and tell them that is wrong that they sell that many units in short length (1,3,5 … meter)

No

I usually have some sort of hypothesis in mind, like if I do that there would be some kind of value adding. I don’'t think there will be

Torben

1 Like

I have no problem with the word standard. Your new cable complies with the latest standard but I don’t believe the ZWP bit is mentioned in it. It would appear that 657.B might be better suited to domestic settings than 657.A due to smaller bend radius.

No-one is wrong to sell short lengths (yes, I know you were joking), but the implication of Fosco’s claim is that people would be misguided if they thought ZWP makes any difference at shorter lengths.

The technical term for this is; Chasing Rabbits :stuck_out_tongue_winking_eye:

Might be fun to try, but changing the fibre will make absolutely no difference whatsoever for home Ethernet Switches where you’re using meter(s) lengths of fibre and grey-optics (1310nm or 1550nm). Your SFPs might see small differences in signal attenuation, but that will have absolutely no consequence to the Ethernet bit stream. If people think they can hear a difference in fibre changes, it’s their imagination :laughing:

The different fibre types are really only important for long reach (>2km) scenarios and coloured-optics transmission (i.e. Wavelength Division Multiplexing over C/L-band frequencies), where the fibre type causes not just signal attenuation, but also non-linear impairments that actually affect the optical signal modulation, and thus the transmission bit error rate.

@Anupc - Ok, let’s make a deal :slight_smile:

I will stop Chasing Rabbits :slight_smile: if you provide me with minimum 2 concrete suggestions that will make an improvement (SFP, FMC, fiber optic …).

If you don’t accept the deal, I will tell everybody that the above cable is HIGHLY recommended from your side due to HUGH (VERY BIG) SQ improvements (LOL - maybe I am just kidding - but you never know with Danes living abroad) :slight_smile:

Torben

I’m sure you’re joking of course😜, but just to be sure, you don’t actually believe you need to “improve” something along that specific chain, do you?

That part of my post was not a joke :slight_smile:

Galvanic isolation from the mains - electrical isolation - Decoupling the power supply from the mains did bring an improvement.

Torben

1 Like

Ahh. Well, I meant what more do you think you need to do (since you’re asking for more suggestions)? And perhaps more importantly, what gives you the impression that something more can/needs to be done? :thinking:

1 Like

When we’re on this exploratory journey, it might be helpful if we could clarify the domain in which we think we can make a difference. Is it digital (jitter and stuff) or analogue (RFI, EMI and other noise)?
Unless I’m mistaken, any optical cable is incapable of conducting noise so a change in sound quality would have to be in the digital domain. Others can argue about ground plane noise and whatever, but changing the optical cable won’t/can’t impact sound quality unless the cable we change to is a broken cable which transmits corrupt digital data. Am I wrong?

1 Like

RFI noise, absolutely

Quite right Nigel.

Once you’re isolated with fibre and you’ve verified packet data transmission integrity, there’s really nothing else to be done on the digital side. So, while it might be a fun learning exercise to play around with SFPs and Fibre, it does absolutely nothing to “improve” the digital domain.

Some folks argue that the digital data is in fact represented by analog signals over the wire/fibre, hence receptive to “improvement”.

While thats true, and even an optical transmission signal is in fact a modulated analog waveform, it’s abstracted with error-correction to pure digital data. By design, physical layer changes, such as changing the SFP or the Fibre, will have zero impact on the abstracted data.

2 Likes